An in silico investigation into the causes of telomere length heterogeneity and its implications for the Hayflick limit

In telomerase-negative cell populations the mean telomere length (TL) decreases with increasing population doubling number (PD). A critically small TL is believed to stop cell proliferation at a cell-, age- and species-specific PD thus defining the Hayflick limit. However, positively skewed TL distributions are broad compared to differences between initial and final mean TL and strongly overlap at middle and late PD, which is inconsistent with a limiting role of TL. We used computer-assisted modelling to define what set of premises may account for the above. Our model incorporates the following concepts. DNA end replication problem: telomeres loose 1 shortening unit (SU) upon each cell division. Free radical-caused TL decrease: telomeres experience random events resulting in the loss of a random SU number within a remaining TL. Stochasticity of gene expression and cell differentiation: cells experience random events inducing mitoses or committing cells to proliferation arrest, the latter option requiring a specified number of mitoses to be passed. Cells whose TL reaches 1 SU cannot divide. The proliferation kinetics of such virtual cells conforms to the transition probability model of cell cycle. When no committing events occur and at realistic SU estimates of the initial TL, maximal PD values far exceed the Hayflick limit observed in normal cells and are consistent with the crisis stage entered by transformed cells that have surpassed the Hayflick limit. At intermediate PD, symmetrical TL distributions are yielded. Upon introduction of committing events making the ratio of the rates of proliferating and committing events (P/C) range from 1.10 to 1.25, TL distributions at intermediate PD become positively skewed, and virtual cell clones show bimodal size distributions. At P/C as high as 1.25 the majority of virtual cells at maximal PD contain telomeres with TL>1 SU. A 10% increase in P/C within the 1.10–1.25 range produces a two-fold increase in the maximal PD, which can reach values of up to 25 observed in rodent and some human cells. Increasing the number of committed mitoses from 0 to 10 can increases PD to about 50 observed in human fibroblasts. Introduction of the random TL breakage makes the shapes of TL distributions quite dissimilar from those observed in real cells. Conclusions: Telomere length decrease is a correlate of cell proliferation that cannot alone account for the Hayflick limit, which primarily depends on parameters of cell population kinetics. Free radical damage influences the Hayflick limit not through TL but rather by affecting the ratio of the rates of events that commit cells to mitoses or to proliferation arrest.